Savitzky-Golay Pre-Processing

step_measure_savitzky_golay creates a specification of a recipe step that smooths and filters the measurement sequence.

Usage

step_measure_savitzky_golay(
  recipe,
  measures = NULL,
  role = NA,
  trained = FALSE,
  degree = 3,
  window_side = 11,
  differentiation_order = 0,
  skip = FALSE,
  id = rand_id("measure_savitzky_golay")
)

Arguments

recipe

A recipe object. The step will be added to the sequence of operations for this recipe.

measures

An optional character vector of measure column names to process. If NULL (the default), all measure columns (columns with class measure_list) will be processed. Use this to limit processing to specific measure columns when working with multiple measurement types.

role

Not used by this step since no new variables are created.

trained

A logical to indicate if the quantities for preprocessing have been estimated.

degree

An integer for the polynomial degree to use for smoothing.

window_side

An integer for how many units there are on each side of the window. This means that window_side = 1 has a total window width of 3 (e.g., width is 2 * window_side + 1).

differentiation_order

An integer for the degree of filtering (zero indicates no differentiation).

skip

A logical. Should the step be skipped when the recipe is baked by recipes::bake()? While all operations are baked when recipes::prep() is run, some operations may not be able to be conducted on new data (e.g. processing the outcome variable(s)). Care should be taken when using skip = TRUE as it may affect the computations for subsequent operations

id

A character string that is unique to this step to identify it.

Value

An updated version of recipe with the new step added to the sequence of any existing operations.

Details

This method can both smooth out random noise and reduce between-predictor correlation. It fits a polynomial to a window of measurements and this results in fewer measurements than the input. Measurements are assumed to be equally spaced.

The polynomial degree should be less than the window size. Also, window size must be greater than polynomial degree. If either case is true, the original argument values are increased to satisfy these conditions (with a warning).

No selectors should be supplied to this step function. The data should be in a special internal format produced by step_measure_input_wide() or step_measure_input_long().

The measurement locations are reset to integer indices starting at one.

Tidying

When you tidy() this step, a tibble with columns is returned.

See also

Other measure-smoothing: step_measure_despike(), step_measure_filter_fourier(), step_measure_smooth_gaussian(), step_measure_smooth_ma(), step_measure_smooth_median(), step_measure_smooth_wavelet()

Examples

if (rlang::is_installed("prospectr")) {
  rec <-
    recipe(water + fat + protein ~ ., data = meats_long) |>
    update_role(id, new_role = "id") |>
    step_measure_input_long(transmittance, location = vars(channel)) |>
    step_measure_savitzky_golay(
      differentiation_order = 1,
      degree = 3,
      window_side = 5
    ) |>
    prep()
}